In [ ]:
import matplotlib.pyplot as plt
import seaborn as sns
# Create a heatmap of the correlation matrix
plt.figure(figsize=(12, 10))
sns.heatmap(correlation, annot=True, cmap='coolwarm', fmt=".2f")
plt.title('Correlation Matrix')
plt.show()
import pandas as pd
data = pd.read_excel('/content/transformed_data (1).xlsx', sheet_name='Sheet1')
# Display the first few rows of the dataset
print("First few rows of the dataset:")
data.head()
First few rows of the dataset:
| Country Name | Year | Exports of goods and services (annual % growth) | Food exports (% of merchandise exports) | Fuel imports (% of merchandise imports) | GDP growth (annual %) | High-technology exports (% of manufactured exports) | Inflation, GDP deflator (annual %) | Inflation, consumer prices (annual %) | Military expenditure (% of GDP) | Trade (% of GDP) | Trade in services (% of GDP) | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | Egypt, Arab Rep. | 2000 | 7.028999 | 7.996953 | 7.584877 | 6.370004 | 0.0 | 3.944407 | 2.683805 | 2.551265 | 39.017936 | 17.344003 |
| 1 | Egypt, Arab Rep. | 2001 | 3.290676 | 9.962074 | 4.924288 | 3.535252 | 0.0 | 1.867700 | 2.269757 | 2.985569 | 39.810427 | 16.630150 |
| 2 | Egypt, Arab Rep. | 2002 | 4.973599 | 8.753934 | 3.975515 | 2.390204 | 0.0 | 3.165579 | 2.737239 | 3.278455 | 40.987068 | 18.731546 |
| 3 | Egypt, Arab Rep. | 2003 | 13.832853 | 8.501811 | 5.168025 | 3.193455 | 0.0 | 6.777494 | 4.507776 | 3.177285 | 46.179641 | 21.854744 |
| 4 | Egypt, Arab Rep. | 2004 | 25.316456 | 9.769097 | 8.572553 | 4.092072 | 0.0 | 11.669908 | 11.270619 | 2.877531 | 57.819905 | 28.199961 |
# Check the columns present in the dataset
print("\nColumns in the dataset:")
print(data.columns)
Columns in the dataset:
Index(['Country Name', 'Year',
'Exports of goods and services (annual % growth)',
'Food exports (% of merchandise exports)',
'Fuel imports (% of merchandise imports)', 'GDP growth (annual %)',
'High-technology exports (% of manufactured exports)',
'Inflation, GDP deflator (annual %)',
'Inflation, consumer prices (annual %)',
'Military expenditure (% of GDP)', 'Trade (% of GDP)',
'Trade in services (% of GDP)'],
dtype='object')
# Check the data types of the columns
print("\nData types of the columns:")
print(data.dtypes)
Data types of the columns: Country Name object Year int64 Exports of goods and services (annual % growth) float64 Food exports (% of merchandise exports) float64 Fuel imports (% of merchandise imports) float64 GDP growth (annual %) float64 High-technology exports (% of manufactured exports) float64 Inflation, GDP deflator (annual %) float64 Inflation, consumer prices (annual %) float64 Military expenditure (% of GDP) float64 Trade (% of GDP) float64 Trade in services (% of GDP) float64 dtype: object
# Check for missing values in the dataset
print("\nMissing values in the dataset:")
print(data.isnull().sum())
Missing values in the dataset: Country Name 0 Year 0 Exports of goods and services (annual % growth) 0 Food exports (% of merchandise exports) 0 Fuel imports (% of merchandise imports) 0 GDP growth (annual %) 0 High-technology exports (% of manufactured exports) 0 Inflation, GDP deflator (annual %) 0 Inflation, consumer prices (annual %) 0 Military expenditure (% of GDP) 0 Trade (% of GDP) 0 Trade in services (% of GDP) 0 dtype: int64
column_means = data[data.columns[2:]].mean()
# Display the mean of each column
print("Mean of each column:")
print(column_means)
Mean of each column: Exports of goods and services (annual % growth) 3.027259 Food exports (% of merchandise exports) 6.186416 Fuel imports (% of merchandise imports) 7.486914 GDP growth (annual %) 2.905349 High-technology exports (% of manufactured exports) 1.749638 Inflation, GDP deflator (annual %) 8.945287 Inflation, consumer prices (annual %) 6.084513 Military expenditure (% of GDP) 3.469588 Trade (% of GDP) 68.498214 Trade in services (% of GDP) 13.379571 dtype: float64
df = data
# Convert columns to numeric
cols_to_convert = df.columns[2:]
df[cols_to_convert] = df[cols_to_convert].apply(pd.to_numeric, errors='coerce')
# Drop rows with any NaN values
df.dropna(inplace=True)
# Print the first few rows of the DataFrame
print(df.head())
Country Name Year Exports of goods and services (annual % growth) \ 0 Egypt, Arab Rep. 2000 7.028999 1 Egypt, Arab Rep. 2001 3.290676 2 Egypt, Arab Rep. 2002 4.973599 3 Egypt, Arab Rep. 2003 13.832853 4 Egypt, Arab Rep. 2004 25.316456 Food exports (% of merchandise exports) \ 0 7.996953 1 9.962074 2 8.753934 3 8.501811 4 9.769097 Fuel imports (% of merchandise imports) GDP growth (annual %) \ 0 7.584877 6.370004 1 4.924288 3.535252 2 3.975515 2.390204 3 5.168025 3.193455 4 8.572553 4.092072 High-technology exports (% of manufactured exports) \ 0 0.0 1 0.0 2 0.0 3 0.0 4 0.0 Inflation, GDP deflator (annual %) Inflation, consumer prices (annual %) \ 0 3.944407 2.683805 1 1.867700 2.269757 2 3.165579 2.737239 3 6.777494 4.507776 4 11.669908 11.270619 Military expenditure (% of GDP) Trade (% of GDP) \ 0 2.551265 39.017936 1 2.985569 39.810427 2 3.278455 40.987068 3 3.177285 46.179641 4 2.877531 57.819905 Trade in services (% of GDP) 0 17.344003 1 16.630150 2 18.731546 3 21.854744 4 28.199961
# Exclude non-numeric columns
numeric_df = df.select_dtypes(include='number')
# Calculate correlation
correlation = numeric_df.corr()
# Print correlation matrix
print(correlation)
Year \
Year 1.000000
Exports of goods and services (annual % growth) -0.057414
Food exports (% of merchandise exports) 0.059655
Fuel imports (% of merchandise imports) -0.039049
GDP growth (annual %) -0.168757
High-technology exports (% of manufactured expo... 0.267271
Inflation, GDP deflator (annual %) 0.051627
Inflation, consumer prices (annual %) 0.089499
Military expenditure (% of GDP) -0.271181
Trade (% of GDP) -0.377001
Trade in services (% of GDP) 0.001444
Exports of goods and services (annual % growth) \
Year -0.057414
Exports of goods and services (annual % growth) 1.000000
Food exports (% of merchandise exports) -0.009530
Fuel imports (% of merchandise imports) 0.074986
GDP growth (annual %) 0.666746
High-technology exports (% of manufactured expo... -0.024633
Inflation, GDP deflator (annual %) -0.028720
Inflation, consumer prices (annual %) -0.013883
Military expenditure (% of GDP) -0.017319
Trade (% of GDP) 0.098266
Trade in services (% of GDP) 0.003992
Food exports (% of merchandise exports) \
Year 0.059655
Exports of goods and services (annual % growth) -0.009530
Food exports (% of merchandise exports) 1.000000
Fuel imports (% of merchandise imports) 0.546381
GDP growth (annual %) -0.122148
High-technology exports (% of manufactured expo... 0.026871
Inflation, GDP deflator (annual %) 0.075261
Inflation, consumer prices (annual %) 0.112819
Military expenditure (% of GDP) -0.160428
Trade (% of GDP) -0.130517
Trade in services (% of GDP) 0.200942
Fuel imports (% of merchandise imports) \
Year -0.039049
Exports of goods and services (annual % growth) 0.074986
Food exports (% of merchandise exports) 0.546381
Fuel imports (% of merchandise imports) 1.000000
GDP growth (annual %) -0.029291
High-technology exports (% of manufactured expo... 0.186013
Inflation, GDP deflator (annual %) -0.017866
Inflation, consumer prices (annual %) 0.161518
Military expenditure (% of GDP) -0.070252
Trade (% of GDP) 0.012419
Trade in services (% of GDP) 0.399689
GDP growth (annual %) \
Year -0.168757
Exports of goods and services (annual % growth) 0.666746
Food exports (% of merchandise exports) -0.122148
Fuel imports (% of merchandise imports) -0.029291
GDP growth (annual %) 1.000000
High-technology exports (% of manufactured expo... -0.012234
Inflation, GDP deflator (annual %) -0.028594
Inflation, consumer prices (annual %) -0.063057
Military expenditure (% of GDP) 0.047418
Trade (% of GDP) 0.120653
Trade in services (% of GDP) -0.017465
High-technology exports (% of manufactured exports) \
Year 0.267271
Exports of goods and services (annual % growth) -0.024633
Food exports (% of merchandise exports) 0.026871
Fuel imports (% of merchandise imports) 0.186013
GDP growth (annual %) -0.012234
High-technology exports (% of manufactured expo... 1.000000
Inflation, GDP deflator (annual %) -0.092827
Inflation, consumer prices (annual %) 0.059484
Military expenditure (% of GDP) 0.116086
Trade (% of GDP) -0.016595
Trade in services (% of GDP) 0.182898
Inflation, GDP deflator (annual %) \
Year 0.051627
Exports of goods and services (annual % growth) -0.028720
Food exports (% of merchandise exports) 0.075261
Fuel imports (% of merchandise imports) -0.017866
GDP growth (annual %) -0.028594
High-technology exports (% of manufactured expo... -0.092827
Inflation, GDP deflator (annual %) 1.000000
Inflation, consumer prices (annual %) 0.483409
Military expenditure (% of GDP) -0.140031
Trade (% of GDP) -0.119524
Trade in services (% of GDP) -0.090090
Inflation, consumer prices (annual %) \
Year 0.089499
Exports of goods and services (annual % growth) -0.013883
Food exports (% of merchandise exports) 0.112819
Fuel imports (% of merchandise imports) 0.161518
GDP growth (annual %) -0.063057
High-technology exports (% of manufactured expo... 0.059484
Inflation, GDP deflator (annual %) 0.483409
Inflation, consumer prices (annual %) 1.000000
Military expenditure (% of GDP) -0.125875
Trade (% of GDP) -0.081537
Trade in services (% of GDP) 0.113212
Military expenditure (% of GDP) \
Year -0.271181
Exports of goods and services (annual % growth) -0.017319
Food exports (% of merchandise exports) -0.160428
Fuel imports (% of merchandise imports) -0.070252
GDP growth (annual %) 0.047418
High-technology exports (% of manufactured expo... 0.116086
Inflation, GDP deflator (annual %) -0.140031
Inflation, consumer prices (annual %) -0.125875
Military expenditure (% of GDP) 1.000000
Trade (% of GDP) 0.214235
Trade in services (% of GDP) 0.212825
Trade (% of GDP) \
Year -0.377001
Exports of goods and services (annual % growth) 0.098266
Food exports (% of merchandise exports) -0.130517
Fuel imports (% of merchandise imports) 0.012419
GDP growth (annual %) 0.120653
High-technology exports (% of manufactured expo... -0.016595
Inflation, GDP deflator (annual %) -0.119524
Inflation, consumer prices (annual %) -0.081537
Military expenditure (% of GDP) 0.214235
Trade (% of GDP) 1.000000
Trade in services (% of GDP) 0.102303
Trade in services (% of GDP)
Year 0.001444
Exports of goods and services (annual % growth) 0.003992
Food exports (% of merchandise exports) 0.200942
Fuel imports (% of merchandise imports) 0.399689
GDP growth (annual %) -0.017465
High-technology exports (% of manufactured expo... 0.182898
Inflation, GDP deflator (annual %) -0.090090
Inflation, consumer prices (annual %) 0.113212
Military expenditure (% of GDP) 0.212825
Trade (% of GDP) 0.102303
Trade in services (% of GDP) 1.000000
import matplotlib.pyplot as plt
import seaborn as sns
# Create a heatmap of the correlation matrix
plt.figure(figsize=(12, 10))
sns.heatmap(correlation, annot=True, cmap='coolwarm', fmt=".2f")
plt.title('Correlation Matrix')
plt.show()
sns.set_style("whitegrid")
# Group data by country
grouped_data = df.groupby('Country Name')
# Plot each variable for each country
for country, data in grouped_data:
plt.figure(figsize=(12, 6))
plt.title(f"{country} - Economic Indicators Over Years")
for column in data.columns[2:]:
plt.plot(data['Year'], data[column], label=column)
plt.xlabel('Year')
plt.ylabel('Value')
plt.legend()
plt.show()
plt.figure(figsize=(12, 8))
sns.boxplot(data=df.iloc[:, 2:])
plt.xticks(rotation=45)
plt.title('Boxplot of Economic Indicators')
plt.show()
sns.pairplot(df, vars=df.columns[2:], diag_kind='kde')
plt.suptitle('Pairplot of Economic Indicators', y=1.02)
plt.show()
cluster countries based on their economic indicators,
data.head()
| Country Name | Year | Exports of goods and services (annual % growth) | Food exports (% of merchandise exports) | Fuel imports (% of merchandise imports) | GDP growth (annual %) | High-technology exports (% of manufactured exports) | Inflation, GDP deflator (annual %) | Inflation, consumer prices (annual %) | Military expenditure (% of GDP) | Trade (% of GDP) | Trade in services (% of GDP) | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 260 | Yemen, Rep. | 2000 | 26.204317 | 0.000000 | 0.000000 | 6.181916 | 0.0 | 23.346052 | 4.590000 | 4.915392 | 75.438839 | 0.0 |
| 261 | Yemen, Rep. | 2001 | -10.013758 | 3.771346 | 12.024777 | 3.803646 | 0.0 | 2.748200 | 11.911591 | 5.481015 | 70.892398 | 0.0 |
| 262 | Yemen, Rep. | 2002 | 7.132448 | 5.178558 | 14.644732 | 3.935232 | 0.0 | 8.711880 | 12.238534 | 6.895608 | 74.730456 | 0.0 |
| 263 | Yemen, Rep. | 2003 | 2.171143 | 4.985725 | 14.864025 | 3.747398 | 0.0 | 10.892343 | 10.832361 | 6.854552 | 74.382617 | 0.0 |
| 264 | Yemen, Rep. | 2004 | 3.811450 | 4.770815 | 12.141815 | 3.972696 | 0.0 | 14.113320 | 12.515095 | 5.301366 | 71.846674 | 0.0 |
from sklearn.cluster import KMeans
import pandas as pd
# Assuming df contains the loaded data from the file
# Grouping the data by 'Country Name' and calculating the mean GDP growth
grouped_data = df.groupby('Country Name')['GDP growth (annual %)'].mean().reset_index()
# Dropping any rows with missing values
grouped_data.dropna(inplace=True)
# Extracting the country names and GDP growth values
countries = grouped_data['Country Name']
gdp_growth = grouped_data['GDP growth (annual %)']
# Reshape the GDP growth values into a 2D array
gdp_growth_2d = gdp_growth.values.reshape(-1, 1)
# Clustering the countries based on GDP growth using KMeans
kmeans = KMeans(n_clusters=3, random_state=0)
clusters = kmeans.fit_predict(gdp_growth_2d)
# Create a DataFrame with countries and their corresponding clusters
country_clusters_df = pd.DataFrame({'Country': countries, 'Cluster': clusters})
# Print the country and its cluster
for cluster_num in range(3):
countries_in_cluster = country_clusters_df[country_clusters_df['Cluster'] == cluster_num]['Country'].unique()
print(f"Cluster {cluster_num}: {', '.join(countries_in_cluster)}")
Cluster 0: Lebanon, Syrian Arab Republic, Yemen, Rep. Cluster 1: Egypt, Arab Rep., Iran, Islamic Rep., Iraq, Israel, Kuwait, Oman, Saudi Arabia, United Arab Emirates, West Bank and Gaza Cluster 2: Libya, Qatar
/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning warnings.warn(
# Plotting the clusters
plt.figure(figsize=(12, 6))
colors = ['blue', 'red', 'green']
for cluster_num in range(3):
countries_in_cluster = country_clusters_df[country_clusters_df['Cluster'] == cluster_num]
plt.scatter(countries_in_cluster['Country'], grouped_data[grouped_data['Country Name'].isin(countries_in_cluster['Country'])]['GDP growth (annual %)'],
color=colors[cluster_num], label=f'Cluster {cluster_num}')
plt.title('Clusters of Countries based on GDP Growth')
plt.xlabel('Country')
plt.ylabel('GDP Growth (%)')
plt.xticks(rotation=90)
plt.legend()
plt.tight_layout()
plt.show()
# Group data by country and calculate the average value for 'Exports of goods and services (annual % growth)'
avg_exports_growth = df.groupby('Country Name')['Exports of goods and services (annual % growth)'].mean().reset_index()
# Perform clustering
from sklearn.cluster import KMeans
# Select the feature for clustering
X = avg_exports_growth['Exports of goods and services (annual % growth)'].values.reshape(-1, 1)
# Define the number of clusters
n_clusters = 3
# Fit KMeans clustering
kmeans = KMeans(n_clusters=n_clusters, random_state=42)
clusters = kmeans.fit_predict(X)
# Add the cluster labels to the DataFrame
avg_exports_growth['Cluster'] = clusters
# Display the countries in each cluster
for cluster_num in range(n_clusters):
countries_in_cluster = avg_exports_growth[avg_exports_growth['Cluster'] == cluster_num]['Country Name'].values
print(f"Cluster {cluster_num}: {', '.join(countries_in_cluster)}")
Cluster 0: Iran, Islamic Rep., Kuwait, Lebanon, Oman, Qatar, Syrian Arab Republic, Yemen, Rep. Cluster 1: Egypt, Arab Rep., Libya Cluster 2: Iraq, Israel, Saudi Arabia, United Arab Emirates, West Bank and Gaza
/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning warnings.warn(
# Plotting the clusters
plt.figure(figsize=(12, 6))
colors = ['blue', 'red', 'green']
for cluster_num in range(n_clusters):
countries_in_cluster = avg_exports_growth[avg_exports_growth['Cluster'] == cluster_num]
plt.scatter(countries_in_cluster['Country Name'], countries_in_cluster['Exports of goods and services (annual % growth)'],
color=colors[cluster_num], label=f'Cluster {cluster_num}')
plt.title('Clusters of Countries based on Exports Growth')
plt.xlabel('Country')
plt.ylabel('Exports Growth (%)')
plt.xticks(rotation=90)
plt.legend()
plt.tight_layout()
plt.show()
india_data = pd.read_excel('transformed_data (2).xlsx')
west_asian_data = pd.read_excel('transformed_data (1).xlsx')
# Merge the two datasets based on the year column
merged_data = pd.merge(india_data, west_asian_data, on='Year', suffixes=('_India', '_WestAsia'))
# Perform regression analysis
from sklearn.linear_model import LinearRegression
X = merged_data[[ 'Exports of goods and services (annual % growth)',
'Food exports (% of merchandise exports)_WestAsia',
'Fuel imports (% of merchandise imports)_WestAsia',
'GDP growth (annual %)_WestAsia',
'High-technology exports (% of manufactured exports)_WestAsia',
'Inflation, GDP deflator (annual %)_WestAsia',
'Inflation, consumer prices (annual %)_WestAsia',
'Military expenditure (% of GDP)_WestAsia', 'Trade (% of GDP)_WestAsia',
'Trade in services (% of GDP)_WestAsia']]
india_data.columns
Index(['Country Name', 'Year', 'Exports of goods and services (% of GDP)',
'Food exports (% of merchandise exports)',
'Fuel imports (% of merchandise imports)', 'GDP growth (annual %)',
'High-technology exports (% of manufactured exports)',
'Inflation, GDP deflator (annual %)',
'Inflation, consumer prices (annual %)',
'Military expenditure (% of GDP)', 'Trade (% of GDP)',
'Trade in services (% of GDP)'],
dtype='object')
from sklearn.preprocessing import StandardScaler
import pandas as pd
# Assuming merged_data is your DataFrame and the columns are as listed
features = ['Exports of goods and services (% of GDP)',
'Fuel imports (% of merchandise imports)_India',
'GDP growth (annual %)_India',
'Inflation, GDP deflator (annual %)_India',
'Inflation, consumer prices (annual %)_India']
# Extract the features into a new DataFrame
data_to_score = merged_data[features]
# Standardize the features
scaler = StandardScaler()
data_standardized = scaler.fit_transform(data_to_score)
# Create a composite score by taking the mean across the standardized features
# This assumes equal weighting of features
merged_data['composite_score'] = data_standardized.mean(axis=1)
# Now merged_data has a new column 'composite_score' which is the standardized average of the selected features
# Define your custom weights for each feature
weights = {
'Exports of goods and services (% of GDP)': 0.2,
'Fuel imports (% of merchandise imports)_India': 0.2,
'GDP growth (annual %)_India': 0.3,
'Inflation, GDP deflator (annual %)_India': 0.15,
'Inflation, consumer prices (annual %)_India': 0.15
}
# Apply weights to each column and sum to get the weighted score
merged_data['weighted_score'] = sum(data_standardized[:, i] * weights[feature] for i, feature in enumerate(features))
# Now merged_data has a new column 'weighted_score' which is the weighted average of the selected features
y = merged_data['weighted_score']
merged_data.columns
Index(['Country Name_India', 'Year',
'Exports of goods and services (% of GDP)',
'Food exports (% of merchandise exports)_India',
'Fuel imports (% of merchandise imports)_India',
'GDP growth (annual %)_India',
'High-technology exports (% of manufactured exports)_India',
'Inflation, GDP deflator (annual %)_India',
'Inflation, consumer prices (annual %)_India',
'Military expenditure (% of GDP)_India', 'Trade (% of GDP)_India',
'Trade in services (% of GDP)_India', 'Country Name_WestAsia',
'Exports of goods and services (annual % growth)',
'Food exports (% of merchandise exports)_WestAsia',
'Fuel imports (% of merchandise imports)_WestAsia',
'GDP growth (annual %)_WestAsia',
'High-technology exports (% of manufactured exports)_WestAsia',
'Inflation, GDP deflator (annual %)_WestAsia',
'Inflation, consumer prices (annual %)_WestAsia',
'Military expenditure (% of GDP)_WestAsia', 'Trade (% of GDP)_WestAsia',
'Trade in services (% of GDP)_WestAsia', 'composite_score',
'weighted_score'],
dtype='object')
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import r2_score, mean_squared_error
# Split the data into training and test sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
# Fit the model
model = LinearRegression()
model.fit(X_train, y_train)
# Make predictions
y_pred = model.predict(X_test)
# Calculate R-squared and RMSE
r2 = r2_score(y_test, y_pred)
rmse = mean_squared_error(y_test, y_pred, squared=False)
print("R-squared:", r2)
print("RMSE:", rmse)
# Plot the predictions against the actual values
plt.figure(figsize=(10, 6))
sns.scatterplot(x=y_test, y=y_pred)
plt.xlabel('Actual GDP growth (annual %)_India')
plt.ylabel('Predicted GDP growth (annual %)_India')
plt.title('Actual vs. Predicted GDP growth (annual %)_India')
plt.show()
# Get model summary
print("Coefficients:", model.coef_)
print("Intercept:", model.intercept_)
R-squared: 0.04464054699629516 RMSE: 0.5979636602836522
Coefficients: [-0.00262082 0.00043487 0.00270506 0.00051244 -0.02773105 -0.01789175 0.00812982 -0.00433155 0.00267284 -0.00010062] Intercept: -0.028594864479177244
The Linear Regression model demonstrates a modest fit with an R-squared of 0.0446 and a RMSE of 0.598, indicating room for improvement in predictive accuracy. The R^2 Score of 0.1221 on the entire dataset suggests that the model explains a small portion of the variance in the data.
# Calculate R^2 score
r_squared = model.score(X, y)
print("R^2 Score:", r_squared)
R^2 Score: 0.12210827866227236
import numpy as np
import pandas as pd
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import LSTM, Dense, Bidirectional
# Assuming X_train, y_train, X_test, y_test are your prepared datasets
# Reshape the data into 3D arrays (samples, time steps, features)
X_train = np.reshape(X_train, (X_train.shape[0], 1, X_train.shape[1]))
X_test = np.reshape(X_test, (X_test.shape[0], 1, X_test.shape[1]))
# Build LSTM model
model_lstm = Sequential()
model_lstm.add(LSTM(50, input_shape=(X_train.shape[1], X_train.shape[2])))
model_lstm.add(Dense(1))
model_lstm.compile(optimizer='adam', loss='mse')
# Build BiLSTM model
model_bilstm = Sequential()
model_bilstm.add(Bidirectional(LSTM(50, input_shape=(X_train.shape[1], X_train.shape[2]))))
model_bilstm.add(Dense(1))
model_bilstm.compile(optimizer='adam', loss='mse')
# Train LSTM model
history_lstm = model_lstm.fit(X_train, y_train, epochs=100, batch_size=32, validation_data=(X_test, y_test), verbose=1)
# Train BiLSTM model
history_bilstm = model_bilstm.fit(X_train, y_train, epochs=100, batch_size=32, validation_data=(X_test, y_test), verbose=1)
# Evaluate LSTM model
lstm_loss = model_lstm.evaluate(X_test, y_test, verbose=0)
print("LSTM Loss:", lstm_loss)
# Evaluate BiLSTM model
bilstm_loss = model_bilstm.evaluate(X_test, y_test, verbose=0)
print("BiLSTM Loss:", bilstm_loss)
# Make predictions using LSTM and BiLSTM models
y_pred_lstm = model_lstm.predict(X_test)
y_pred_bilstm = model_bilstm.predict(X_test)
# Plot predictions vs. actual
plt.figure(figsize=(10, 6))
plt.plot(y_test, label='Actual')
plt.plot(y_pred_lstm, label='LSTM Predictions')
plt.plot(y_pred_bilstm, label='BiLSTM Predictions')
plt.legend()
plt.show()
Epoch 1/100 7/7 [==============================] - 3s 102ms/step - loss: 0.9265 - val_loss: 0.5617 Epoch 2/100 7/7 [==============================] - 0s 8ms/step - loss: 0.7490 - val_loss: 0.4566 Epoch 3/100 7/7 [==============================] - 0s 11ms/step - loss: 0.6735 - val_loss: 0.4206 Epoch 4/100 7/7 [==============================] - 0s 8ms/step - loss: 0.6314 - val_loss: 0.4007 Epoch 5/100 7/7 [==============================] - 0s 9ms/step - loss: 0.5974 - val_loss: 0.3845 Epoch 6/100 7/7 [==============================] - 0s 12ms/step - loss: 0.5674 - val_loss: 0.3659 Epoch 7/100 7/7 [==============================] - 0s 10ms/step - loss: 0.5395 - val_loss: 0.3571 Epoch 8/100 7/7 [==============================] - 0s 9ms/step - loss: 0.5159 - val_loss: 0.3507 Epoch 9/100 7/7 [==============================] - 0s 9ms/step - loss: 0.4974 - val_loss: 0.3433 Epoch 10/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4821 - val_loss: 0.3417 Epoch 11/100 7/7 [==============================] - 0s 9ms/step - loss: 0.4679 - val_loss: 0.3454 Epoch 12/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4525 - val_loss: 0.3418 Epoch 13/100 7/7 [==============================] - 0s 9ms/step - loss: 0.4416 - val_loss: 0.3378 Epoch 14/100 7/7 [==============================] - 0s 9ms/step - loss: 0.4319 - val_loss: 0.3350 Epoch 15/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4196 - val_loss: 0.3423 Epoch 16/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4061 - val_loss: 0.3312 Epoch 17/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3922 - val_loss: 0.3272 Epoch 18/100 7/7 [==============================] - 0s 8ms/step - loss: 0.3832 - val_loss: 0.3268 Epoch 19/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3717 - val_loss: 0.3321 Epoch 20/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3617 - val_loss: 0.3329 Epoch 21/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3566 - val_loss: 0.3362 Epoch 22/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3466 - val_loss: 0.3336 Epoch 23/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3401 - val_loss: 0.3329 Epoch 24/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3322 - val_loss: 0.3384 Epoch 25/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3242 - val_loss: 0.3373 Epoch 26/100 7/7 [==============================] - 0s 8ms/step - loss: 0.3192 - val_loss: 0.3392 Epoch 27/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3127 - val_loss: 0.3421 Epoch 28/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3051 - val_loss: 0.3674 Epoch 29/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3026 - val_loss: 0.3585 Epoch 30/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2983 - val_loss: 0.3407 Epoch 31/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2885 - val_loss: 0.3515 Epoch 32/100 7/7 [==============================] - 0s 8ms/step - loss: 0.2810 - val_loss: 0.3453 Epoch 33/100 7/7 [==============================] - 0s 9ms/step - loss: 0.2784 - val_loss: 0.3422 Epoch 34/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2768 - val_loss: 0.3374 Epoch 35/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2667 - val_loss: 0.3760 Epoch 36/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2640 - val_loss: 0.3664 Epoch 37/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2558 - val_loss: 0.3624 Epoch 38/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2503 - val_loss: 0.3704 Epoch 39/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2478 - val_loss: 0.3667 Epoch 40/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2377 - val_loss: 0.4081 Epoch 41/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2379 - val_loss: 0.4040 Epoch 42/100 7/7 [==============================] - 0s 8ms/step - loss: 0.2321 - val_loss: 0.3917 Epoch 43/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2224 - val_loss: 0.4017 Epoch 44/100 7/7 [==============================] - 0s 9ms/step - loss: 0.2166 - val_loss: 0.3910 Epoch 45/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2125 - val_loss: 0.3897 Epoch 46/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2084 - val_loss: 0.3939 Epoch 47/100 7/7 [==============================] - 0s 9ms/step - loss: 0.2038 - val_loss: 0.4008 Epoch 48/100 7/7 [==============================] - 0s 8ms/step - loss: 0.2007 - val_loss: 0.4011 Epoch 49/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1957 - val_loss: 0.3875 Epoch 50/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1943 - val_loss: 0.4225 Epoch 51/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1885 - val_loss: 0.4156 Epoch 52/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1848 - val_loss: 0.3902 Epoch 53/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1819 - val_loss: 0.3955 Epoch 54/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1782 - val_loss: 0.3998 Epoch 55/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1727 - val_loss: 0.4024 Epoch 56/100 7/7 [==============================] - 0s 8ms/step - loss: 0.1689 - val_loss: 0.4049 Epoch 57/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1649 - val_loss: 0.4256 Epoch 58/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1623 - val_loss: 0.4037 Epoch 59/100 7/7 [==============================] - 0s 8ms/step - loss: 0.1581 - val_loss: 0.4004 Epoch 60/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1547 - val_loss: 0.4056 Epoch 61/100 7/7 [==============================] - 0s 8ms/step - loss: 0.1521 - val_loss: 0.4035 Epoch 62/100 7/7 [==============================] - 0s 8ms/step - loss: 0.1489 - val_loss: 0.4044 Epoch 63/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1461 - val_loss: 0.4246 Epoch 64/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1501 - val_loss: 0.4138 Epoch 65/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1426 - val_loss: 0.4643 Epoch 66/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1453 - val_loss: 0.4400 Epoch 67/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1365 - val_loss: 0.4272 Epoch 68/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1337 - val_loss: 0.4389 Epoch 69/100 7/7 [==============================] - 0s 8ms/step - loss: 0.1328 - val_loss: 0.4567 Epoch 70/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1294 - val_loss: 0.4546 Epoch 71/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1267 - val_loss: 0.4485 Epoch 72/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1252 - val_loss: 0.4454 Epoch 73/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1231 - val_loss: 0.4429 Epoch 74/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1201 - val_loss: 0.4669 Epoch 75/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1189 - val_loss: 0.4555 Epoch 76/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1178 - val_loss: 0.4623 Epoch 77/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1137 - val_loss: 0.4981 Epoch 78/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1188 - val_loss: 0.4746 Epoch 79/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1123 - val_loss: 0.4970 Epoch 80/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1124 - val_loss: 0.5134 Epoch 81/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1114 - val_loss: 0.4749 Epoch 82/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1101 - val_loss: 0.5156 Epoch 83/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1052 - val_loss: 0.4837 Epoch 84/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1034 - val_loss: 0.4958 Epoch 85/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1030 - val_loss: 0.4894 Epoch 86/100 7/7 [==============================] - 0s 11ms/step - loss: 0.0989 - val_loss: 0.5112 Epoch 87/100 7/7 [==============================] - 0s 9ms/step - loss: 0.0997 - val_loss: 0.5186 Epoch 88/100 7/7 [==============================] - 0s 11ms/step - loss: 0.0979 - val_loss: 0.5011 Epoch 89/100 7/7 [==============================] - 0s 9ms/step - loss: 0.0968 - val_loss: 0.5116 Epoch 90/100 7/7 [==============================] - 0s 12ms/step - loss: 0.0963 - val_loss: 0.5098 Epoch 91/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0922 - val_loss: 0.5151 Epoch 92/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0906 - val_loss: 0.5159 Epoch 93/100 7/7 [==============================] - 0s 11ms/step - loss: 0.0952 - val_loss: 0.5396 Epoch 94/100 7/7 [==============================] - 0s 9ms/step - loss: 0.0958 - val_loss: 0.5438 Epoch 95/100 7/7 [==============================] - 0s 11ms/step - loss: 0.0894 - val_loss: 0.5274 Epoch 96/100 7/7 [==============================] - 0s 9ms/step - loss: 0.0880 - val_loss: 0.5510 Epoch 97/100 7/7 [==============================] - 0s 8ms/step - loss: 0.0877 - val_loss: 0.5463 Epoch 98/100 7/7 [==============================] - 0s 11ms/step - loss: 0.0879 - val_loss: 0.5252 Epoch 99/100 7/7 [==============================] - 0s 9ms/step - loss: 0.0841 - val_loss: 0.5570 Epoch 100/100 7/7 [==============================] - 0s 11ms/step - loss: 0.0820 - val_loss: 0.5376 Epoch 1/100 7/7 [==============================] - 6s 157ms/step - loss: 0.9102 - val_loss: 0.4865 Epoch 2/100 7/7 [==============================] - 0s 12ms/step - loss: 0.7198 - val_loss: 0.4059 Epoch 3/100 7/7 [==============================] - 0s 12ms/step - loss: 0.6747 - val_loss: 0.3776 Epoch 4/100 7/7 [==============================] - 0s 10ms/step - loss: 0.6316 - val_loss: 0.3615 Epoch 5/100 7/7 [==============================] - 0s 13ms/step - loss: 0.5838 - val_loss: 0.3468 Epoch 6/100 7/7 [==============================] - 0s 10ms/step - loss: 0.5569 - val_loss: 0.3353 Epoch 7/100 7/7 [==============================] - 0s 13ms/step - loss: 0.5266 - val_loss: 0.3135 Epoch 8/100 7/7 [==============================] - 0s 12ms/step - loss: 0.5015 - val_loss: 0.3022 Epoch 9/100 7/7 [==============================] - 0s 10ms/step - loss: 0.4822 - val_loss: 0.2901 Epoch 10/100 7/7 [==============================] - 0s 10ms/step - loss: 0.4640 - val_loss: 0.2825 Epoch 11/100 7/7 [==============================] - 0s 10ms/step - loss: 0.4469 - val_loss: 0.2710 Epoch 12/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4308 - val_loss: 0.2649 Epoch 13/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4165 - val_loss: 0.2562 Epoch 14/100 7/7 [==============================] - 0s 13ms/step - loss: 0.4079 - val_loss: 0.2557 Epoch 15/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3954 - val_loss: 0.2529 Epoch 16/100 7/7 [==============================] - 0s 13ms/step - loss: 0.3856 - val_loss: 0.2534 Epoch 17/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3763 - val_loss: 0.2541 Epoch 18/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3646 - val_loss: 0.2673 Epoch 19/100 7/7 [==============================] - 0s 13ms/step - loss: 0.3577 - val_loss: 0.2660 Epoch 20/100 7/7 [==============================] - 0s 13ms/step - loss: 0.3508 - val_loss: 0.2568 Epoch 21/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3400 - val_loss: 0.2696 Epoch 22/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3365 - val_loss: 0.2751 Epoch 23/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3264 - val_loss: 0.2632 Epoch 24/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3177 - val_loss: 0.2678 Epoch 25/100 7/7 [==============================] - 0s 13ms/step - loss: 0.3116 - val_loss: 0.2674 Epoch 26/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3039 - val_loss: 0.2647 Epoch 27/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2962 - val_loss: 0.2731 Epoch 28/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2893 - val_loss: 0.2734 Epoch 29/100 7/7 [==============================] - 0s 13ms/step - loss: 0.2854 - val_loss: 0.2806 Epoch 30/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2779 - val_loss: 0.2746 Epoch 31/100 7/7 [==============================] - 0s 13ms/step - loss: 0.2710 - val_loss: 0.2716 Epoch 32/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2651 - val_loss: 0.2840 Epoch 33/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2561 - val_loss: 0.2806 Epoch 34/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2557 - val_loss: 0.2800 Epoch 35/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2481 - val_loss: 0.3286 Epoch 36/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2387 - val_loss: 0.3389 Epoch 37/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2365 - val_loss: 0.3550 Epoch 38/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2281 - val_loss: 0.3355 Epoch 39/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2222 - val_loss: 0.3440 Epoch 40/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2123 - val_loss: 0.3259 Epoch 41/100 7/7 [==============================] - 0s 13ms/step - loss: 0.2113 - val_loss: 0.3299 Epoch 42/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2049 - val_loss: 0.3437 Epoch 43/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2008 - val_loss: 0.3632 Epoch 44/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1957 - val_loss: 0.3370 Epoch 45/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1923 - val_loss: 0.3812 Epoch 46/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1886 - val_loss: 0.3897 Epoch 47/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1853 - val_loss: 0.3775 Epoch 48/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1776 - val_loss: 0.3777 Epoch 49/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1771 - val_loss: 0.3795 Epoch 50/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1711 - val_loss: 0.3693 Epoch 51/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1745 - val_loss: 0.3816 Epoch 52/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1645 - val_loss: 0.4110 Epoch 53/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1637 - val_loss: 0.3995 Epoch 54/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1603 - val_loss: 0.4007 Epoch 55/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1547 - val_loss: 0.4182 Epoch 56/100 7/7 [==============================] - 0s 15ms/step - loss: 0.1558 - val_loss: 0.4047 Epoch 57/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1542 - val_loss: 0.3931 Epoch 58/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1526 - val_loss: 0.4131 Epoch 59/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1531 - val_loss: 0.3906 Epoch 60/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1471 - val_loss: 0.4260 Epoch 61/100 7/7 [==============================] - 0s 16ms/step - loss: 0.1433 - val_loss: 0.3991 Epoch 62/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1401 - val_loss: 0.4098 Epoch 63/100 7/7 [==============================] - 0s 16ms/step - loss: 0.1336 - val_loss: 0.4404 Epoch 64/100 7/7 [==============================] - 0s 16ms/step - loss: 0.1338 - val_loss: 0.4341 Epoch 65/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1314 - val_loss: 0.4481 Epoch 66/100 7/7 [==============================] - 0s 18ms/step - loss: 0.1288 - val_loss: 0.4421 Epoch 67/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1248 - val_loss: 0.4340 Epoch 68/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1273 - val_loss: 0.4491 Epoch 69/100 7/7 [==============================] - 0s 15ms/step - loss: 0.1374 - val_loss: 0.4488 Epoch 70/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1240 - val_loss: 0.4215 Epoch 71/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1225 - val_loss: 0.4765 Epoch 72/100 7/7 [==============================] - 0s 16ms/step - loss: 0.1217 - val_loss: 0.4167 Epoch 73/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1241 - val_loss: 0.4226 Epoch 74/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1114 - val_loss: 0.4700 Epoch 75/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1113 - val_loss: 0.4533 Epoch 76/100 7/7 [==============================] - 0s 16ms/step - loss: 0.1091 - val_loss: 0.4657 Epoch 77/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1061 - val_loss: 0.4805 Epoch 78/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1053 - val_loss: 0.4565 Epoch 79/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1040 - val_loss: 0.4760 Epoch 80/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1037 - val_loss: 0.4809 Epoch 81/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1046 - val_loss: 0.4631 Epoch 82/100 7/7 [==============================] - 0s 17ms/step - loss: 0.0997 - val_loss: 0.4737 Epoch 83/100 7/7 [==============================] - 0s 16ms/step - loss: 0.0982 - val_loss: 0.4427 Epoch 84/100 7/7 [==============================] - 0s 16ms/step - loss: 0.0955 - val_loss: 0.4906 Epoch 85/100 7/7 [==============================] - 0s 20ms/step - loss: 0.0982 - val_loss: 0.4727 Epoch 86/100 7/7 [==============================] - 0s 14ms/step - loss: 0.0924 - val_loss: 0.4955 Epoch 87/100 7/7 [==============================] - 0s 12ms/step - loss: 0.0921 - val_loss: 0.4627 Epoch 88/100 7/7 [==============================] - 0s 12ms/step - loss: 0.0890 - val_loss: 0.4835 Epoch 89/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0872 - val_loss: 0.4782 Epoch 90/100 7/7 [==============================] - 0s 12ms/step - loss: 0.0895 - val_loss: 0.4835 Epoch 91/100 7/7 [==============================] - 0s 13ms/step - loss: 0.0856 - val_loss: 0.4997 Epoch 92/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0841 - val_loss: 0.5006 Epoch 93/100 7/7 [==============================] - 0s 12ms/step - loss: 0.0837 - val_loss: 0.4960 Epoch 94/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0822 - val_loss: 0.5085 Epoch 95/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0822 - val_loss: 0.5090 Epoch 96/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0793 - val_loss: 0.4935 Epoch 97/100 7/7 [==============================] - 0s 13ms/step - loss: 0.0772 - val_loss: 0.4940 Epoch 98/100 7/7 [==============================] - 0s 12ms/step - loss: 0.0772 - val_loss: 0.5195 Epoch 99/100 7/7 [==============================] - 0s 13ms/step - loss: 0.0761 - val_loss: 0.5015 Epoch 100/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0747 - val_loss: 0.4994 LSTM Loss: 0.5376214981079102 BiLSTM Loss: 0.49937736988067627 2/2 [==============================] - 0s 5ms/step 2/2 [==============================] - 1s 6ms/step
from sklearn.metrics import mean_squared_error, r2_score, mean_absolute_error
def calculate_regression_metrics(y_true, y_pred):
"""
Calculate and return various regression metrics.
"""
mse = mean_squared_error(y_true, y_pred)
rmse = mean_squared_error(y_true, y_pred, squared=False)
mae = mean_absolute_error(y_true, y_pred)
r2 = r2_score(y_true, y_pred)
return {
'Mean Squared Error (MSE)': mse,
'Root Mean Squared Error (RMSE)': rmse,
'Mean Absolute Error (MAE)': mae,
'R-squared': r2
}
def plot_actual_vs_predicted(y_true, y_pred):
"""
Plot the actual vs. predicted values.
"""
plt.figure(figsize=(10, 6))
sns.scatterplot(x=y_true, y=y_pred)
plt.xlabel('Actual')
plt.ylabel('Predicted')
plt.title('Actual vs. Predicted')
plt.show()
# Example usage
def evaluate_model(y_true, y_pred):
metrics = calculate_regression_metrics(y_true, y_pred)
print("LSTM Metrics:")
for metric, value in metrics.items():
print(f"{metric}: {value}")
plot_actual_vs_predicted(y_true, y_pred)
y_pred_lstm_flat = y_pred_lstm.flatten()
# Assuming y_true is the actual values and y_pred_lstm_flat is the flattened predicted values from your LSTM model
evaluate_model(y_test, y_pred_lstm_flat)
LSTM Metrics: Mean Squared Error (MSE): 0.5376215368280253 Root Mean Squared Error (RMSE): 0.7332267976745157 Mean Absolute Error (MAE): 0.5642939971376032 R-squared: -0.4364611339803315
# Modify the evaluate_model function to return the metrics
def evaluate_model(y_true, y_pred):
metrics = calculate_regression_metrics(y_true, y_pred)
print("Regression Metrics:")
for metric, value in metrics.items():
print(f"{metric}: {value}")
plot_actual_vs_predicted(y_true, y_pred)
return metrics
# Use the function to get the metrics
metrics = evaluate_model(y_test, y_pred_lstm_flat)
# Convert the metrics to a DataFrame
metrics_df = pd.DataFrame(list(metrics.items()), columns=['Metric', 'Value'])
# Create the barplot
plt.figure(figsize=(15, 6))
sns.barplot(x='Metric', y='Value', data=metrics_df)
plt.title('Regression Metrics')
plt.show()
Regression Metrics: Mean Squared Error (MSE): 0.5376215368280253 Root Mean Squared Error (RMSE): 0.7332267976745157 Mean Absolute Error (MAE): 0.5642939971376032 R-squared: -0.4364611339803315
from tensorflow.keras.layers import GRU
# Build GRU model
model_gru = Sequential()
model_gru.add(GRU(50, input_shape=(X_train.shape[1], X_train.shape[2])))
model_gru.add(Dense(1))
model_gru.compile(optimizer='adam', loss='mse')
# Build BiGRU model
model_bigru = Sequential()
model_bigru.add(Bidirectional(GRU(50, input_shape=(X_train.shape[1], X_train.shape[2]))))
model_bigru.add(Dense(1))
model_bigru.compile(optimizer='adam', loss='mse')
# Train GRU model
history_gru = model_gru.fit(X_train, y_train, epochs=100, batch_size=32, validation_data=(X_test, y_test), verbose=1)
# Train BiGRU model
history_bigru = model_bigru.fit(X_train, y_train, epochs=100, batch_size=32, validation_data=(X_test, y_test), verbose=1)
# Evaluate GRU model
gru_loss = model_gru.evaluate(X_test, y_test, verbose=0)
print("GRU Loss:", gru_loss)
# Evaluate BiGRU model
bigru_loss = model_bigru.evaluate(X_test, y_test, verbose=0)
print("BiGRU Loss:", bigru_loss)
# Make predictions using GRU and BiGRU models
y_pred_gru = model_gru.predict(X_test)
y_pred_bigru = model_bigru.predict(X_test)
# Plot predictions vs. actual
plt.figure(figsize=(15, 6))
plt.plot(y_test, label='Actual')
plt.plot(y_pred_gru, label='GRU Predictions')
plt.plot(y_pred_bigru, label='BiGRU Predictions')
plt.legend()
plt.show()
Epoch 1/100 7/7 [==============================] - 3s 89ms/step - loss: 1.5857 - val_loss: 0.7272 Epoch 2/100 7/7 [==============================] - 0s 9ms/step - loss: 0.9762 - val_loss: 0.5158 Epoch 3/100 7/7 [==============================] - 0s 11ms/step - loss: 0.7661 - val_loss: 0.4915 Epoch 4/100 7/7 [==============================] - 0s 9ms/step - loss: 0.6872 - val_loss: 0.4811 Epoch 5/100 7/7 [==============================] - 0s 10ms/step - loss: 0.6478 - val_loss: 0.4377 Epoch 6/100 7/7 [==============================] - 0s 9ms/step - loss: 0.6036 - val_loss: 0.3939 Epoch 7/100 7/7 [==============================] - 0s 9ms/step - loss: 0.5651 - val_loss: 0.3630 Epoch 8/100 7/7 [==============================] - 0s 11ms/step - loss: 0.5359 - val_loss: 0.3439 Epoch 9/100 7/7 [==============================] - 0s 11ms/step - loss: 0.5108 - val_loss: 0.3330 Epoch 10/100 7/7 [==============================] - 0s 9ms/step - loss: 0.4892 - val_loss: 0.3274 Epoch 11/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4706 - val_loss: 0.3253 Epoch 12/100 7/7 [==============================] - 0s 10ms/step - loss: 0.4524 - val_loss: 0.3349 Epoch 13/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4441 - val_loss: 0.3357 Epoch 14/100 7/7 [==============================] - 0s 9ms/step - loss: 0.4294 - val_loss: 0.3247 Epoch 15/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4207 - val_loss: 0.3164 Epoch 16/100 7/7 [==============================] - 0s 12ms/step - loss: 0.4128 - val_loss: 0.3089 Epoch 17/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4025 - val_loss: 0.3166 Epoch 18/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3943 - val_loss: 0.3165 Epoch 19/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3860 - val_loss: 0.3106 Epoch 20/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3798 - val_loss: 0.3040 Epoch 21/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3729 - val_loss: 0.3023 Epoch 22/100 7/7 [==============================] - 0s 8ms/step - loss: 0.3664 - val_loss: 0.3088 Epoch 23/100 7/7 [==============================] - 0s 8ms/step - loss: 0.3615 - val_loss: 0.3042 Epoch 24/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3551 - val_loss: 0.2981 Epoch 25/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3482 - val_loss: 0.3063 Epoch 26/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3442 - val_loss: 0.3138 Epoch 27/100 7/7 [==============================] - 0s 8ms/step - loss: 0.3392 - val_loss: 0.3140 Epoch 28/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3330 - val_loss: 0.3185 Epoch 29/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3285 - val_loss: 0.3191 Epoch 30/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3251 - val_loss: 0.3205 Epoch 31/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3194 - val_loss: 0.3129 Epoch 32/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3149 - val_loss: 0.3006 Epoch 33/100 7/7 [==============================] - 0s 11ms/step - loss: 0.3117 - val_loss: 0.3024 Epoch 34/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3067 - val_loss: 0.2923 Epoch 35/100 7/7 [==============================] - 0s 9ms/step - loss: 0.3049 - val_loss: 0.2964 Epoch 36/100 7/7 [==============================] - 0s 9ms/step - loss: 0.2978 - val_loss: 0.3139 Epoch 37/100 7/7 [==============================] - 0s 9ms/step - loss: 0.2962 - val_loss: 0.3201 Epoch 38/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2912 - val_loss: 0.3083 Epoch 39/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2859 - val_loss: 0.3116 Epoch 40/100 7/7 [==============================] - 0s 16ms/step - loss: 0.2852 - val_loss: 0.3120 Epoch 41/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2849 - val_loss: 0.3322 Epoch 42/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2758 - val_loss: 0.3104 Epoch 43/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2727 - val_loss: 0.3029 Epoch 44/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2701 - val_loss: 0.2969 Epoch 45/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2650 - val_loss: 0.2990 Epoch 46/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2633 - val_loss: 0.3023 Epoch 47/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2615 - val_loss: 0.2951 Epoch 48/100 7/7 [==============================] - 0s 13ms/step - loss: 0.2563 - val_loss: 0.3090 Epoch 49/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2536 - val_loss: 0.3149 Epoch 50/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2497 - val_loss: 0.3202 Epoch 51/100 7/7 [==============================] - 0s 17ms/step - loss: 0.2483 - val_loss: 0.3117 Epoch 52/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2442 - val_loss: 0.3143 Epoch 53/100 7/7 [==============================] - 0s 13ms/step - loss: 0.2413 - val_loss: 0.3029 Epoch 54/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2386 - val_loss: 0.3100 Epoch 55/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2392 - val_loss: 0.3143 Epoch 56/100 7/7 [==============================] - 0s 13ms/step - loss: 0.2354 - val_loss: 0.3264 Epoch 57/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2309 - val_loss: 0.3056 Epoch 58/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2275 - val_loss: 0.3117 Epoch 59/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2249 - val_loss: 0.3155 Epoch 60/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2251 - val_loss: 0.3249 Epoch 61/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2224 - val_loss: 0.3059 Epoch 62/100 7/7 [==============================] - 0s 17ms/step - loss: 0.2175 - val_loss: 0.3152 Epoch 63/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2158 - val_loss: 0.3220 Epoch 64/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2129 - val_loss: 0.3043 Epoch 65/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2111 - val_loss: 0.3246 Epoch 66/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2112 - val_loss: 0.3433 Epoch 67/100 7/7 [==============================] - 0s 8ms/step - loss: 0.2050 - val_loss: 0.3133 Epoch 68/100 7/7 [==============================] - 0s 9ms/step - loss: 0.2030 - val_loss: 0.3165 Epoch 69/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2010 - val_loss: 0.3301 Epoch 70/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1983 - val_loss: 0.3202 Epoch 71/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1936 - val_loss: 0.3105 Epoch 72/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1943 - val_loss: 0.3117 Epoch 73/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1902 - val_loss: 0.3368 Epoch 74/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1921 - val_loss: 0.3296 Epoch 75/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1866 - val_loss: 0.3372 Epoch 76/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1857 - val_loss: 0.3290 Epoch 77/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1828 - val_loss: 0.3300 Epoch 78/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1817 - val_loss: 0.3356 Epoch 79/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1790 - val_loss: 0.3367 Epoch 80/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1777 - val_loss: 0.3378 Epoch 81/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1739 - val_loss: 0.3417 Epoch 82/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1740 - val_loss: 0.3408 Epoch 83/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1736 - val_loss: 0.3269 Epoch 84/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1742 - val_loss: 0.3834 Epoch 85/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1713 - val_loss: 0.3597 Epoch 86/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1673 - val_loss: 0.3411 Epoch 87/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1676 - val_loss: 0.3474 Epoch 88/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1619 - val_loss: 0.3248 Epoch 89/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1618 - val_loss: 0.3278 Epoch 90/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1581 - val_loss: 0.3470 Epoch 91/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1582 - val_loss: 0.3486 Epoch 92/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1600 - val_loss: 0.3454 Epoch 93/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1534 - val_loss: 0.3528 Epoch 94/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1521 - val_loss: 0.3459 Epoch 95/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1519 - val_loss: 0.3618 Epoch 96/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1497 - val_loss: 0.3520 Epoch 97/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1487 - val_loss: 0.3471 Epoch 98/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1483 - val_loss: 0.3504 Epoch 99/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1465 - val_loss: 0.3596 Epoch 100/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1444 - val_loss: 0.3694 Epoch 1/100 7/7 [==============================] - 5s 148ms/step - loss: 2.0028 - val_loss: 0.5859 Epoch 2/100 7/7 [==============================] - 0s 10ms/step - loss: 0.9073 - val_loss: 0.5746 Epoch 3/100 7/7 [==============================] - 0s 10ms/step - loss: 0.8248 - val_loss: 0.4978 Epoch 4/100 7/7 [==============================] - 0s 10ms/step - loss: 0.6905 - val_loss: 0.3378 Epoch 5/100 7/7 [==============================] - 0s 9ms/step - loss: 0.6073 - val_loss: 0.3116 Epoch 6/100 7/7 [==============================] - 0s 12ms/step - loss: 0.5759 - val_loss: 0.2963 Epoch 7/100 7/7 [==============================] - 0s 12ms/step - loss: 0.5336 - val_loss: 0.2951 Epoch 8/100 7/7 [==============================] - 0s 12ms/step - loss: 0.5037 - val_loss: 0.2983 Epoch 9/100 7/7 [==============================] - 0s 10ms/step - loss: 0.4777 - val_loss: 0.2829 Epoch 10/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4542 - val_loss: 0.2769 Epoch 11/100 7/7 [==============================] - 0s 13ms/step - loss: 0.4343 - val_loss: 0.2682 Epoch 12/100 7/7 [==============================] - 0s 11ms/step - loss: 0.4190 - val_loss: 0.2592 Epoch 13/100 7/7 [==============================] - 0s 12ms/step - loss: 0.4030 - val_loss: 0.2629 Epoch 14/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3862 - val_loss: 0.2749 Epoch 15/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3730 - val_loss: 0.2809 Epoch 16/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3602 - val_loss: 0.2836 Epoch 17/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3472 - val_loss: 0.2791 Epoch 18/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3375 - val_loss: 0.2797 Epoch 19/100 7/7 [==============================] - 0s 12ms/step - loss: 0.3259 - val_loss: 0.2805 Epoch 20/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3169 - val_loss: 0.2812 Epoch 21/100 7/7 [==============================] - 0s 13ms/step - loss: 0.3082 - val_loss: 0.2988 Epoch 22/100 7/7 [==============================] - 0s 10ms/step - loss: 0.3046 - val_loss: 0.3006 Epoch 23/100 7/7 [==============================] - 0s 12ms/step - loss: 0.2934 - val_loss: 0.2870 Epoch 24/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2876 - val_loss: 0.2838 Epoch 25/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2810 - val_loss: 0.2935 Epoch 26/100 7/7 [==============================] - 0s 11ms/step - loss: 0.2756 - val_loss: 0.2954 Epoch 27/100 7/7 [==============================] - 0s 10ms/step - loss: 0.2692 - val_loss: 0.2781 Epoch 28/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2649 - val_loss: 0.2887 Epoch 29/100 7/7 [==============================] - 0s 16ms/step - loss: 0.2594 - val_loss: 0.3010 Epoch 30/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2560 - val_loss: 0.2873 Epoch 31/100 7/7 [==============================] - 0s 17ms/step - loss: 0.2506 - val_loss: 0.3133 Epoch 32/100 7/7 [==============================] - 0s 16ms/step - loss: 0.2487 - val_loss: 0.3205 Epoch 33/100 7/7 [==============================] - 0s 19ms/step - loss: 0.2423 - val_loss: 0.2996 Epoch 34/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2405 - val_loss: 0.3106 Epoch 35/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2333 - val_loss: 0.3115 Epoch 36/100 7/7 [==============================] - 0s 18ms/step - loss: 0.2301 - val_loss: 0.3160 Epoch 37/100 7/7 [==============================] - 0s 14ms/step - loss: 0.2268 - val_loss: 0.3255 Epoch 38/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2291 - val_loss: 0.3446 Epoch 39/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2187 - val_loss: 0.3198 Epoch 40/100 7/7 [==============================] - 0s 15ms/step - loss: 0.2179 - val_loss: 0.3359 Epoch 41/100 7/7 [==============================] - 0s 16ms/step - loss: 0.2122 - val_loss: 0.3188 Epoch 42/100 7/7 [==============================] - 0s 17ms/step - loss: 0.2096 - val_loss: 0.3269 Epoch 43/100 7/7 [==============================] - 0s 17ms/step - loss: 0.2017 - val_loss: 0.3501 Epoch 44/100 7/7 [==============================] - 0s 18ms/step - loss: 0.2017 - val_loss: 0.3399 Epoch 45/100 7/7 [==============================] - 0s 16ms/step - loss: 0.1998 - val_loss: 0.3616 Epoch 46/100 7/7 [==============================] - 0s 19ms/step - loss: 0.2017 - val_loss: 0.3373 Epoch 47/100 7/7 [==============================] - 0s 15ms/step - loss: 0.1920 - val_loss: 0.3841 Epoch 48/100 7/7 [==============================] - 0s 18ms/step - loss: 0.1898 - val_loss: 0.3464 Epoch 49/100 7/7 [==============================] - 0s 15ms/step - loss: 0.1839 - val_loss: 0.3558 Epoch 50/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1810 - val_loss: 0.3523 Epoch 51/100 7/7 [==============================] - 0s 18ms/step - loss: 0.1776 - val_loss: 0.3399 Epoch 52/100 7/7 [==============================] - 0s 17ms/step - loss: 0.1749 - val_loss: 0.3512 Epoch 53/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1733 - val_loss: 0.3541 Epoch 54/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1695 - val_loss: 0.3581 Epoch 55/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1667 - val_loss: 0.3567 Epoch 56/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1664 - val_loss: 0.3537 Epoch 57/100 7/7 [==============================] - 0s 9ms/step - loss: 0.1600 - val_loss: 0.3879 Epoch 58/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1594 - val_loss: 0.3545 Epoch 59/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1633 - val_loss: 0.3435 Epoch 60/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1535 - val_loss: 0.3908 Epoch 61/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1523 - val_loss: 0.3762 Epoch 62/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1513 - val_loss: 0.3556 Epoch 63/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1501 - val_loss: 0.3867 Epoch 64/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1457 - val_loss: 0.4200 Epoch 65/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1459 - val_loss: 0.3950 Epoch 66/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1417 - val_loss: 0.4208 Epoch 67/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1394 - val_loss: 0.3903 Epoch 68/100 7/7 [==============================] - 0s 15ms/step - loss: 0.1365 - val_loss: 0.4049 Epoch 69/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1364 - val_loss: 0.4120 Epoch 70/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1409 - val_loss: 0.4414 Epoch 71/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1324 - val_loss: 0.3949 Epoch 72/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1325 - val_loss: 0.4378 Epoch 73/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1321 - val_loss: 0.4377 Epoch 74/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1272 - val_loss: 0.4049 Epoch 75/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1254 - val_loss: 0.4147 Epoch 76/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1273 - val_loss: 0.4198 Epoch 77/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1223 - val_loss: 0.4190 Epoch 78/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1215 - val_loss: 0.4398 Epoch 79/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1170 - val_loss: 0.4389 Epoch 80/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1173 - val_loss: 0.4123 Epoch 81/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1172 - val_loss: 0.4472 Epoch 82/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1171 - val_loss: 0.4490 Epoch 83/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1138 - val_loss: 0.4354 Epoch 84/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1136 - val_loss: 0.4400 Epoch 85/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1141 - val_loss: 0.4414 Epoch 86/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1084 - val_loss: 0.4338 Epoch 87/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1139 - val_loss: 0.4577 Epoch 88/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1095 - val_loss: 0.4786 Epoch 89/100 7/7 [==============================] - 0s 11ms/step - loss: 0.1116 - val_loss: 0.4513 Epoch 90/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1073 - val_loss: 0.4935 Epoch 91/100 7/7 [==============================] - 0s 12ms/step - loss: 0.1082 - val_loss: 0.4809 Epoch 92/100 7/7 [==============================] - 0s 15ms/step - loss: 0.1064 - val_loss: 0.4559 Epoch 93/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1030 - val_loss: 0.4684 Epoch 94/100 7/7 [==============================] - 0s 10ms/step - loss: 0.1030 - val_loss: 0.4498 Epoch 95/100 7/7 [==============================] - 0s 14ms/step - loss: 0.1021 - val_loss: 0.5100 Epoch 96/100 7/7 [==============================] - 0s 13ms/step - loss: 0.1001 - val_loss: 0.4570 Epoch 97/100 7/7 [==============================] - 0s 13ms/step - loss: 0.0999 - val_loss: 0.4987 Epoch 98/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0991 - val_loss: 0.4855 Epoch 99/100 7/7 [==============================] - 0s 10ms/step - loss: 0.0988 - val_loss: 0.4716 Epoch 100/100 7/7 [==============================] - 0s 13ms/step - loss: 0.0962 - val_loss: 0.4940 GRU Loss: 0.3694315254688263 BiGRU Loss: 0.49399012327194214 2/2 [==============================] - 0s 6ms/step 2/2 [==============================] - 1s 9ms/step
y_pred_bilstm_flat = y_pred_bilstm.flatten()
print("BiLSTM METRICS")
# Use the function to get the metrics
metrics = evaluate_model(y_test, y_pred_bilstm_flat)
# Convert the metrics to a DataFrame
metrics_df = pd.DataFrame(list(metrics.items()), columns=['Metric', 'Value'])
# Create the barplot
plt.figure(figsize=(10, 6))
sns.barplot(x='Metric', y='Value', data=metrics_df)
plt.title('Regression Metrics')
plt.show()
BiLSTM METRICS Regression Metrics: Mean Squared Error (MSE): 0.4993773689000158 Root Mean Squared Error (RMSE): 0.7066663773663041 Mean Absolute Error (MAE): 0.518508944506656 R-squared: -0.33427724240089884
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
y_pred_gru_flat = y_pred_gru.flatten()
print("GRU METRICS")
# Use the function to get the metrics
metrics = evaluate_model(y_test, y_pred_gru_flat)
# Convert the metrics to a DataFrame
metrics_df = pd.DataFrame(list(metrics.items()), columns=['Metric', 'Value'])
# Create the barplot
plt.figure(figsize=(10, 6))
sns.barplot(x='Metric', y='Value', data=metrics_df)
plt.title('Regression Metrics')
plt.show()
GRU METRICS Regression Metrics: Mean Squared Error (MSE): 0.3694315387835335 Root Mean Squared Error (RMSE): 0.6078088011731432 Mean Absolute Error (MAE): 0.4875229776845975 R-squared: 0.01292264022740286
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
y_pred_bigru_flat = y_pred_bigru.flatten()
print("BiGRU METRICS")
# Use the function to get the metrics
metrics = evaluate_model(y_test, y_pred_bigru_flat)
# Convert the metrics to a DataFrame
metrics_df = pd.DataFrame(list(metrics.items()), columns=['Metric', 'Value'])
# Create the barplot
plt.figure(figsize=(10, 6))
sns.barplot(x='Metric', y='Value', data=metrics_df)
plt.title('Regression Metrics')
plt.show()
BiGRU METRICS Regression Metrics: Mean Squared Error (MSE): 0.4939901358552205 Root Mean Squared Error (RMSE): 0.7028443183630501 Mean Absolute Error (MAE): 0.5503988972015131 R-squared: -0.31988319313308033
import pandas as pd
import matplotlib.pyplot as plt
# Create a DataFrame with the metrics
metrics_data = {
'Model': ['LSTM', 'BiLSTM', 'GRU', 'BiGRU'],
'MSE': [0.4562067420626733, 0.46446107029808337, 0.41529824621674843, 0.5616803301187765],
'RMSE': [0.6754307825844728, 0.6815138078557788, 0.644436378719225, 0.7494533541980959],
'MAE': [0.5365917077435299, 0.5579181404310574, 0.4737416878096157, 0.5797752033972754],
'R-squared': [-0.21893043552391278, -0.24098502390078402, -0.1096277749962633, -0.5007433829700807]
}
df_metrics = pd.DataFrame(metrics_data)
# Plotting the metrics
fig, axs = plt.subplots(2, 2, figsize=(15, 10))
# MSE plot
axs[0, 0].bar(df_metrics['Model'], df_metrics['MSE'], color='skyblue')
axs[0, 0].set_title('Mean Squared Error (MSE)')
axs[0, 0].set_ylabel('MSE')
# RMSE plot
axs[0, 1].bar(df_metrics['Model'], df_metrics['RMSE'], color='lightgreen')
axs[0, 1].set_title('Root Mean Squared Error (RMSE)')
axs[0, 1].set_ylabel('RMSE')
# MAE plot
axs[1, 0].bar(df_metrics['Model'], df_metrics['MAE'], color='salmon')
axs[1, 0].set_title('Mean Absolute Error (MAE)')
axs[1, 0].set_ylabel('MAE')
# R-squared plot
axs[1, 1].bar(df_metrics['Model'], df_metrics['R-squared'], color='orchid')
axs[1, 1].set_title('R-squared')
axs[1, 1].set_ylabel('R-squared')
plt.tight_layout()
plt.show()
Based on the provided metrics for each model, here are the inferences and conclusions:
LSTM Model:
The LSTM model has a moderate level of errors as indicated by the MSE, RMSE, and MAE. However, the negative R-squared value suggests that the model is not capturing the underlying pattern of the data and is performing worse than a simple horizontal line fit.
BiLSTM Model:
The BiLSTM model shows slightly higher errors compared to the LSTM model. The negative R-squared value is also worse, indicating that the bidirectional nature of the model did not contribute to better performance and may have added complexity without benefit.
GRU Model:
The GRU model has the lowest errors among the models and a less negative R-squared value, suggesting it performed better than both LSTM and BiLSTM models. However, the negative R-squared value still indicates a poor fit to the data.
BiGRU Model:
The BiGRU model has the highest errors and the most negative R-squared value, which implies that it performed the worst among all the models. This suggests that the bidirectional GRU did not capture the sequential nature of the data effectively.
Conclusion: All models have negative R-squared values, which is a strong indicator that they are not suitable for the dataset in their current form. The GRU model shows relatively better performance, but still not a good fit. This could be due to several factors such as inadequate feature selection, need for hyperparameter tuning, or the complexity of the data itself. It is recommended to revisit the data preprocessing steps, consider feature engineering, and possibly explore different modeling approaches or algorithms. Additionally, evaluating the models on different metrics and conducting error analysis could provide further insights into improving model performance.
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import Ridge, Lasso
from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import mean_squared_error, r2_score
import matplotlib.pyplot as plt
import seaborn as sns
# Split the data into training and test sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
# Initialize machine learning models
ridge_model = Ridge(random_state=42)
lasso_model = Lasso(random_state=42)
rf_model = RandomForestRegressor(random_state=42)
gb_model = GradientBoostingRegressor(random_state=42)
# Create a dictionary of models for easy access
models = {'Ridge': ridge_model, 'Lasso': lasso_model, 'Random Forest': rf_model, 'Gradient Boosting': gb_model}
# Train and evaluate each model
model_metrics = {}
for name, model in models.items():
model.fit(X_train, y_train)
y_pred = model.predict(X_test)
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)
model_metrics[name] = {'MSE': mse, 'R-squared': r2}
print(f'{name} - MSE: {mse}, R-squared: {r2}')
# Find the best model based on R-squared
best_model_name = max(model_metrics, key=lambda k: model_metrics[k]['R-squared'])
best_model = models[best_model_name]
y_pred_best = best_model.predict(X_test)
# Plotting the results for the best model
plt.figure(figsize=(10, 6))
sns.scatterplot(x=y_test, y=y_pred_best)
plt.xlabel('Actual Weighted Score')
plt.ylabel('Predicted Weighted Score')
plt.title(f'Actual vs. Predicted Weighted Score for {best_model_name}')
plt.show()
Ridge - MSE: 0.3575427842636131, R-squared: 0.0446879856041108 Lasso - MSE: 0.33736560137653193, R-squared: 0.09859903087493371 Random Forest - MSE: 0.2772359066771713, R-squared: 0.25925846044939305 Gradient Boosting - MSE: 0.26687389134109546, R-squared: 0.286944539373613
# Function to calculate metrics
def calculate_metrics(y_true, y_pred):
mse = mean_squared_error(y_true, y_pred)
rmse = np.sqrt(mse)
mae = mean_absolute_error(y_true, y_pred)
r2 = r2_score(y_true, y_pred)
return mse, rmse, mae, r2
# Calculate metrics for Ridge
y_pred_ridge=ridge_model.predict(X_test)
ridge_mse, ridge_rmse, ridge_mae, ridge_r2 = calculate_metrics(y_test, y_pred_ridge)
# Calculate metrics for Lasso
y_pred_lasso=lasso_model.predict(X_test)
lasso_mse, lasso_rmse, lasso_mae, lasso_r2 = calculate_metrics(y_test, y_pred_lasso)
# Calculate metrics for Random Forest
y_pred_rf=rf_model.predict(X_test)
rf_mse, rf_rmse, rf_mae, rf_r2 = calculate_metrics(y_test, y_pred_rf)
# Calculate metrics for Gradient Boosting
y_pred_gb=gb_model.predict(X_test)
gb_mse, gb_rmse, gb_mae, gb_r2 = calculate_metrics(y_test, y_pred_gb)
# Print the metrics
print(f'Ridge - MSE: {ridge_mse}, RMSE: {ridge_rmse}, MAE: {ridge_mae}, R-squared: {ridge_r2}')
print(f'Lasso - MSE: {lasso_mse}, RMSE: {lasso_rmse}, MAE: {lasso_mae}, R-squared: {lasso_r2}')
print(f'Random Forest - MSE: {rf_mse}, RMSE: {rf_rmse}, MAE: {rf_mae}, R-squared: {rf_r2}')
print(f'Gradient Boosting - MSE: {gb_mse}, RMSE: {gb_rmse}, MAE: {gb_mae}, R-squared: {gb_r2}')
Ridge - MSE: 0.3575427842636131, RMSE: 0.5979488140832901, MAE: 0.47871089062785516, R-squared: 0.0446879856041108 Lasso - MSE: 0.33736560137653193, RMSE: 0.5808318184952783, MAE: 0.451416547604268, R-squared: 0.09859903087493371 Random Forest - MSE: 0.2772359066771713, RMSE: 0.5265319616862506, MAE: 0.3896422695461326, R-squared: 0.25925846044939305 Gradient Boosting - MSE: 0.26687389134109546, RMSE: 0.5165983849578853, MAE: 0.3946339132125805, R-squared: 0.286944539373613
# Existing metrics data
metrics_data = {
'Model': ['LSTM', 'BiLSTM', 'GRU', 'BiGRU', 'Ridge', 'Lasso', 'Random Forest', 'Gradient Boosting'],
'MSE': [0.4562067420626733, 0.46446107029808337, 0.41529824621674843, 0.5616803301187765, 0.3575427842636131, 0.33736560137653193, 0.2772359066771713, 0.26687389134109546],
'RMSE': [0.6754307825844728, 0.6815138078557788, 0.644436378719225, 0.7494533541980959, 0.5979488140832901, 0.5808318184952783, 0.5265319616862506, 0.5165983849578853],
'MAE': [0.5365917077435299, 0.5579181404310574, 0.4737416878096157, 0.5797752033972754, 0.47871089062785516, 0.451416547604268, 0.3896422695461326, 0.3946339132125805],
'R-squared': [-0.21893043552391278, -0.24098502390078402, -0.1096277749962633, -0.5007433829700807, 0.0446879856041108, 0.09859903087493371, 0.25925846044939305, 0.286944539373613]
}
df_metrics = pd.DataFrame(metrics_data)
# Plotting the updated metrics
fig, axs = plt.subplots(2, 2, figsize=(20, 15))
# MSE plot
axs[0, 0].bar(df_metrics['Model'], df_metrics['MSE'], color='skyblue')
axs[0, 0].set_title('Mean Squared Error (MSE)')
axs[0, 0].set_ylabel('MSE')
# RMSE plot
axs[0, 1].bar(df_metrics['Model'], df_metrics['RMSE'], color='lightgreen')
axs[0, 1].set_title('Root Mean Squared Error (RMSE)')
axs[0, 1].set_ylabel('RMSE')
# MAE plot
axs[1, 0].bar(df_metrics['Model'], df_metrics['MAE'], color='salmon')
axs[1, 0].set_title('Mean Absolute Error (MAE)')
axs[1, 0].set_ylabel('MAE')
# R-squared plot
axs[1, 1].bar(df_metrics['Model'], df_metrics['R-squared'], color='orchid')
axs[1, 1].set_title('R-squared')
axs[1, 1].set_ylabel('R-squared')
plt.tight_layout()
plt.show()
conclusions and inferences:
# Feature importance for Random Forest
rf_feature_importance = rf_model.feature_importances_
# Feature importance for Gradient Boosting
gb_feature_importance = gb_model.feature_importances_
# Coefficients for Ridge
ridge_coefficients = ridge_model.coef_
# Coefficients for Lasso
lasso_coefficients = lasso_model.coef_
# Create DataFrame for each model's feature importance/coefficients
df_rf_importance = pd.DataFrame({'Feature': X.columns, 'Importance': rf_feature_importance})
df_gb_importance = pd.DataFrame({'Feature': X.columns, 'Importance': gb_feature_importance})
df_ridge_coefficients = pd.DataFrame({'Feature': X.columns, 'Coefficient': ridge_coefficients})
df_lasso_coefficients = pd.DataFrame({'Feature': X.columns, 'Coefficient': lasso_coefficients})
# Sort DataFrame by importance/coefficient
df_rf_importance = df_rf_importance.sort_values(by='Importance', ascending=False)
df_gb_importance = df_gb_importance.sort_values(by='Importance', ascending=False)
df_ridge_coefficients = df_ridge_coefficients.sort_values(by='Coefficient', ascending=False)
df_lasso_coefficients = df_lasso_coefficients.sort_values(by='Coefficient', ascending=False)
# Print the top 5 most influential variables for each model
print("Random Forest Feature Importance:")
print(df_rf_importance)
print("\nGradient Boosting Feature Importance:")
print(df_gb_importance)
print("\nRidge Coefficients:")
print(df_ridge_coefficients)
print("\nLasso Coefficients:")
print(df_lasso_coefficients)
Random Forest Feature Importance:
Feature Importance
5 Inflation, GDP deflator (annual %)_WestAsia 0.263810
8 Trade (% of GDP)_WestAsia 0.140702
3 GDP growth (annual %)_WestAsia 0.130898
4 High-technology exports (% of manufactured exp... 0.124515
6 Inflation, consumer prices (annual %)_WestAsia 0.086614
0 Exports of goods and services (annual % growth) 0.067723
7 Military expenditure (% of GDP)_WestAsia 0.054579
2 Fuel imports (% of merchandise imports)_WestAsia 0.050327
1 Food exports (% of merchandise exports)_WestAsia 0.047328
9 Trade in services (% of GDP)_WestAsia 0.033505
Gradient Boosting Feature Importance:
Feature Importance
5 Inflation, GDP deflator (annual %)_WestAsia 0.282827
8 Trade (% of GDP)_WestAsia 0.166055
3 GDP growth (annual %)_WestAsia 0.160629
4 High-technology exports (% of manufactured exp... 0.143541
7 Military expenditure (% of GDP)_WestAsia 0.068283
6 Inflation, consumer prices (annual %)_WestAsia 0.058728
0 Exports of goods and services (annual % growth) 0.034823
1 Food exports (% of merchandise exports)_WestAsia 0.030417
9 Trade in services (% of GDP)_WestAsia 0.028838
2 Fuel imports (% of merchandise imports)_WestAsia 0.025860
Ridge Coefficients:
Feature Coefficient
6 Inflation, consumer prices (annual %)_WestAsia 0.008129
2 Fuel imports (% of merchandise imports)_WestAsia 0.002705
8 Trade (% of GDP)_WestAsia 0.002673
3 GDP growth (annual %)_WestAsia 0.000512
1 Food exports (% of merchandise exports)_WestAsia 0.000435
9 Trade in services (% of GDP)_WestAsia -0.000101
0 Exports of goods and services (annual % growth) -0.002621
7 Military expenditure (% of GDP)_WestAsia -0.004330
5 Inflation, GDP deflator (annual %)_WestAsia -0.017891
4 High-technology exports (% of manufactured exp... -0.027723
Lasso Coefficients:
Feature Coefficient
8 Trade (% of GDP)_WestAsia 0.001685
1 Food exports (% of merchandise exports)_WestAsia 0.000000
2 Fuel imports (% of merchandise imports)_WestAsia 0.000000
3 GDP growth (annual %)_WestAsia -0.000000
4 High-technology exports (% of manufactured exp... -0.000000
6 Inflation, consumer prices (annual %)_WestAsia 0.000000
7 Military expenditure (% of GDP)_WestAsia -0.000000
9 Trade in services (% of GDP)_WestAsia 0.000000
0 Exports of goods and services (annual % growth) -0.000161
5 Inflation, GDP deflator (annual %)_WestAsia -0.009962
import matplotlib.pyplot as plt
import seaborn as sns
# Set up the matplotlib figure
plt.figure(figsize=(20, 15))
# Plot Random Forest Feature Importance
plt.subplot(221)
sns.barplot(x='Importance', y='Feature', data=df_rf_importance.head(5), palette='viridis')
plt.title('Random Forest - Top 5 Most Influential Variables')
plt.xlabel('Importance')
plt.ylabel('Feature')
# Plot Gradient Boosting Feature Importance
plt.subplot(222)
sns.barplot(x='Importance', y='Feature', data=df_gb_importance.head(5), palette='viridis')
plt.title('Gradient Boosting - Top 5 Most Influential Variables')
plt.xlabel('Importance')
plt.ylabel('Feature')
# Plot Ridge Coefficients
plt.subplot(223)
sns.barplot(x='Coefficient', y='Feature', data=df_ridge_coefficients.head(5), palette='viridis')
plt.title('Ridge - Top 5 Most Influential Variables')
plt.xlabel('Coefficient')
plt.ylabel('Feature')
# Plot Lasso Coefficients
plt.subplot(224)
sns.barplot(x='Coefficient', y='Feature', data=df_lasso_coefficients.head(5), palette='viridis')
plt.title('Lasso - Top 5 Most Influential Variables')
plt.xlabel('Coefficient')
plt.ylabel('Feature')
# Adjust layout
plt.tight_layout()
<ipython-input-41-69f96f8ae2f0>:9: FutureWarning: Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect. sns.barplot(x='Importance', y='Feature', data=df_rf_importance.head(5), palette='viridis') <ipython-input-41-69f96f8ae2f0>:16: FutureWarning: Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect. sns.barplot(x='Importance', y='Feature', data=df_gb_importance.head(5), palette='viridis') <ipython-input-41-69f96f8ae2f0>:23: FutureWarning: Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect. sns.barplot(x='Coefficient', y='Feature', data=df_ridge_coefficients.head(5), palette='viridis') <ipython-input-41-69f96f8ae2f0>:30: FutureWarning: Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect. sns.barplot(x='Coefficient', y='Feature', data=df_lasso_coefficients.head(5), palette='viridis')
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
# Normalize the importance values to a probability scale (0 to 1) using min-max normalization
def normalize_importance(df):
df['Importance_normalized'] = (df['Importance'] - df['Importance'].min()) / (df['Importance'].max() - df['Importance'].min())
return df
# Assuming df_rf_importance, df_gb_importance, df_ridge_coefficients, and df_lasso_coefficients contain the feature importances or coefficients
# Create normalized dataframes
df_rf_importance_normalized = normalize_importance(df_rf_importance)
df_gb_importance_normalized = normalize_importance(df_gb_importance)
df_ridge_coefficients_normalized = normalize_importance(df_ridge_coefficients)
df_lasso_coefficients_normalized = normalize_importance(df_lasso_coefficients)
# Set up the matplotlib figure
plt.figure(figsize=(20, 15))
# Plot Random Forest Feature Importance
plt.subplot(221)
sns.barplot(x='Importance_normalized', y='Feature', data=df_rf_importance_normalized.head(5), palette='viridis')
plt.title('Random Forest - Top 5 Most Influential Variables')
plt.xlabel('Importance (Normalized)')
plt.ylabel('Feature')
# Plot Gradient Boosting Feature Importance
plt.subplot(222)
sns.barplot(x='Importance_normalized', y='Feature', data=df_gb_importance_normalized.head(5), palette='viridis')
plt.title('Gradient Boosting - Top 5 Most Influential Variables')
plt.xlabel('Importance (Normalized)')
plt.ylabel('Feature')
# Plot Ridge Coefficients
plt.subplot(223)
sns.barplot(x='Importance_normalized', y='Feature', data=df_ridge_coefficients_normalized.head(5), palette='viridis')
plt.title('Ridge - Top 5 Most Influential Variables')
plt.xlabel('Coefficient (Normalized)')
plt.ylabel('Feature')
# Plot Lasso Coefficients
plt.subplot(224)
sns.barplot(x='Importance_normalized', y='Feature', data=df_lasso_coefficients_normalized.head(5), palette='viridis')
plt.title('Lasso - Top 5 Most Influential Variables')
plt.xlabel('Coefficient (Normalized)')
plt.ylabel('Feature')
# Adjust layout
plt.tight_layout()
# Show the plots
plt.show()
--------------------------------------------------------------------------- KeyError Traceback (most recent call last) /usr/local/lib/python3.10/dist-packages/pandas/core/indexes/base.py in get_loc(self, key) 3652 try: -> 3653 return self._engine.get_loc(casted_key) 3654 except KeyError as err: /usr/local/lib/python3.10/dist-packages/pandas/_libs/index.pyx in pandas._libs.index.IndexEngine.get_loc() /usr/local/lib/python3.10/dist-packages/pandas/_libs/index.pyx in pandas._libs.index.IndexEngine.get_loc() pandas/_libs/hashtable_class_helper.pxi in pandas._libs.hashtable.PyObjectHashTable.get_item() pandas/_libs/hashtable_class_helper.pxi in pandas._libs.hashtable.PyObjectHashTable.get_item() KeyError: 'Importance' The above exception was the direct cause of the following exception: KeyError Traceback (most recent call last) <ipython-input-48-d2101d52c8dd> in <cell line: 15>() 13 df_rf_importance_normalized = normalize_importance(df_rf_importance) 14 df_gb_importance_normalized = normalize_importance(df_gb_importance) ---> 15 df_ridge_coefficients_normalized = normalize_importance(df_ridge_coefficients) 16 df_lasso_coefficients_normalized = normalize_importance(df_lasso_coefficients) 17 <ipython-input-48-d2101d52c8dd> in normalize_importance(df) 5 # Normalize the importance values to a probability scale (0 to 1) using min-max normalization 6 def normalize_importance(df): ----> 7 df['Importance_normalized'] = (df['Importance'] - df['Importance'].min()) / (df['Importance'].max() - df['Importance'].min()) 8 return df 9 /usr/local/lib/python3.10/dist-packages/pandas/core/frame.py in __getitem__(self, key) 3759 if self.columns.nlevels > 1: 3760 return self._getitem_multilevel(key) -> 3761 indexer = self.columns.get_loc(key) 3762 if is_integer(indexer): 3763 indexer = [indexer] /usr/local/lib/python3.10/dist-packages/pandas/core/indexes/base.py in get_loc(self, key) 3653 return self._engine.get_loc(casted_key) 3654 except KeyError as err: -> 3655 raise KeyError(key) from err 3656 except TypeError: 3657 # If we have a listlike key, _check_indexing_error will raise KeyError: 'Importance'
import pandas as pd
# Assuming india_data and west_asian_data are your DataFrames
# Group the data by country and calculate the average for each economic indicator
india_grouped = india_data.groupby('Country Name').mean()
west_asian_grouped = west_asian_data.groupby('Country Name').mean()
# Merge the grouped data on 'Country Name'
merged_grouped = pd.merge(india_grouped, west_asian_grouped, on='Country Name', suffixes=('_India', '_WestAsia'))
# Select columns for correlation analysis
columns_to_correlate = [
'Exports of goods and services (% of GDP)_India',
'Food exports (% of merchandise exports)_India',
'Fuel imports (% of merchandise imports)_India',
'GDP growth (annual %)_India',
'High-technology exports (% of manufactured exports)_India',
'Inflation, GDP deflator (annual %)_India',
'Inflation, consumer prices (annual %)_India',
'Military expenditure (% of GDP)_India',
'Trade (% of GDP)_India',
'Trade in services (% of GDP)_India',
'Exports of goods and services (annual % growth)_WestAsia',
'Food exports (% of merchandise exports)_WestAsia',
'Fuel imports (% of merchandise imports)_WestAsia',
'GDP growth (annual %)_WestAsia',
'High-technology exports (% of manufactured exports)_WestAsia',
'Inflation, GDP deflator (annual %)_WestAsia',
'Inflation, consumer prices (annual %)_WestAsia',
'Military expenditure (% of GDP)_WestAsia',
'Trade (% of GDP)_WestAsia',
'Trade in services (% of GDP)_WestAsia'
]
# Calculate correlation matrix
correlation_matrix = merged_grouped[columns_to_correlate].corr()
# Plot the correlation matrix
import matplotlib.pyplot as plt
import seaborn as sns
plt.figure(figsize=(16, 10))
sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', center=0)
plt.title('Correlation Matrix of Economic Indicators')
plt.show()
--------------------------------------------------------------------------- KeyError Traceback (most recent call last) <ipython-input-54-06fbc5d1942e> in <cell line: 36>() 34 35 # Calculate correlation matrix ---> 36 correlation_matrix = merged_grouped[columns_to_correlate].corr() 37 38 # Plot the correlation matrix /usr/local/lib/python3.10/dist-packages/pandas/core/frame.py in __getitem__(self, key) 3765 if is_iterator(key): 3766 key = list(key) -> 3767 indexer = self.columns._get_indexer_strict(key, "columns")[1] 3768 3769 # take() does not accept boolean indexers /usr/local/lib/python3.10/dist-packages/pandas/core/indexes/base.py in _get_indexer_strict(self, key, axis_name) 5875 keyarr, indexer, new_indexer = self._reindex_non_unique(keyarr) 5876 -> 5877 self._raise_if_missing(keyarr, indexer, axis_name) 5878 5879 keyarr = self.take(indexer) /usr/local/lib/python3.10/dist-packages/pandas/core/indexes/base.py in _raise_if_missing(self, key, indexer, axis_name) 5939 5940 not_found = list(ensure_index(key)[missing_mask.nonzero()[0]].unique()) -> 5941 raise KeyError(f"{not_found} not in index") 5942 5943 @overload KeyError: "['Exports of goods and services (% of GDP)_India', 'Exports of goods and services (annual % growth)_WestAsia'] not in index"
west_asian_data.shape
(280, 12)
!pip install nbconvert
Requirement already satisfied: nbconvert in /usr/local/lib/python3.10/dist-packages (6.5.4) Requirement already satisfied: lxml in /usr/local/lib/python3.10/dist-packages (from nbconvert) (4.9.4) Requirement already satisfied: beautifulsoup4 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (4.12.3) Requirement already satisfied: bleach in /usr/local/lib/python3.10/dist-packages (from nbconvert) (6.1.0) Requirement already satisfied: defusedxml in /usr/local/lib/python3.10/dist-packages (from nbconvert) (0.7.1) Requirement already satisfied: entrypoints>=0.2.2 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (0.4) Requirement already satisfied: jinja2>=3.0 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (3.1.3) Requirement already satisfied: jupyter-core>=4.7 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (5.7.2) Requirement already satisfied: jupyterlab-pygments in /usr/local/lib/python3.10/dist-packages (from nbconvert) (0.3.0) Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (2.1.5) Requirement already satisfied: mistune<2,>=0.8.1 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (0.8.4) Requirement already satisfied: nbclient>=0.5.0 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (0.10.0) Requirement already satisfied: nbformat>=5.1 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (5.10.4) Requirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from nbconvert) (24.0) Requirement already satisfied: pandocfilters>=1.4.1 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (1.5.1) Requirement already satisfied: pygments>=2.4.1 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (2.16.1) Requirement already satisfied: tinycss2 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (1.2.1) Requirement already satisfied: traitlets>=5.0 in /usr/local/lib/python3.10/dist-packages (from nbconvert) (5.7.1) Requirement already satisfied: platformdirs>=2.5 in /usr/local/lib/python3.10/dist-packages (from jupyter-core>=4.7->nbconvert) (4.2.0) Requirement already satisfied: jupyter-client>=6.1.12 in /usr/local/lib/python3.10/dist-packages (from nbclient>=0.5.0->nbconvert) (6.1.12) Requirement already satisfied: fastjsonschema>=2.15 in /usr/local/lib/python3.10/dist-packages (from nbformat>=5.1->nbconvert) (2.19.1) Requirement already satisfied: jsonschema>=2.6 in /usr/local/lib/python3.10/dist-packages (from nbformat>=5.1->nbconvert) (4.19.2) Requirement already satisfied: soupsieve>1.2 in /usr/local/lib/python3.10/dist-packages (from beautifulsoup4->nbconvert) (2.5) Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.10/dist-packages (from bleach->nbconvert) (1.16.0) Requirement already satisfied: webencodings in /usr/local/lib/python3.10/dist-packages (from bleach->nbconvert) (0.5.1) Requirement already satisfied: attrs>=22.2.0 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6->nbformat>=5.1->nbconvert) (23.2.0) Requirement already satisfied: jsonschema-specifications>=2023.03.6 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6->nbformat>=5.1->nbconvert) (2023.12.1) Requirement already satisfied: referencing>=0.28.4 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6->nbformat>=5.1->nbconvert) (0.34.0) Requirement already satisfied: rpds-py>=0.7.1 in /usr/local/lib/python3.10/dist-packages (from jsonschema>=2.6->nbformat>=5.1->nbconvert) (0.18.0) Requirement already satisfied: pyzmq>=13 in /usr/local/lib/python3.10/dist-packages (from jupyter-client>=6.1.12->nbclient>=0.5.0->nbconvert) (23.2.1) Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.10/dist-packages (from jupyter-client>=6.1.12->nbclient>=0.5.0->nbconvert) (2.8.2) Requirement already satisfied: tornado>=4.1 in /usr/local/lib/python3.10/dist-packages (from jupyter-client>=6.1.12->nbclient>=0.5.0->nbconvert) (6.3.3)
!apt-get install texlive texlive-xetex texlive-latex-extra pandoc
Reading package lists... Done Building dependency tree... Done Reading state information... Done pandoc is already the newest version (2.9.2.1-3ubuntu2). texlive is already the newest version (2021.20220204-1). texlive-latex-extra is already the newest version (2021.20220204-1). texlive-xetex is already the newest version (2021.20220204-1). 0 upgraded, 0 newly installed, 0 to remove and 45 not upgraded.
!jupyter nbconvert --to pdf Copy_of_fin_analytics_try_2_j_comp.ipynb
[NbConvertApp] Converting notebook Copy_of_fin_analytics_try_2_j_comp.ipynb to pdf
[NbConvertApp] Support files will be in Copy_of_fin_analytics_try_2_j_comp_files/
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Making directory ./Copy_of_fin_analytics_try_2_j_comp_files
[NbConvertApp] Writing 223293 bytes to notebook.tex
[NbConvertApp] Building PDF
[NbConvertApp] Running xelatex 3 times: ['xelatex', 'notebook.tex', '-quiet']
Traceback (most recent call last):
File "/usr/local/bin/jupyter-nbconvert", line 8, in <module>
sys.exit(main())
File "/usr/local/lib/python3.10/dist-packages/jupyter_core/application.py", line 283, in launch_instance
super().launch_instance(argv=argv, **kwargs)
File "/usr/local/lib/python3.10/dist-packages/traitlets/config/application.py", line 992, in launch_instance
app.start()
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 423, in start
self.convert_notebooks()
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 597, in convert_notebooks
self.convert_single_notebook(notebook_filename)
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 560, in convert_single_notebook
output, resources = self.export_single_notebook(
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 488, in export_single_notebook
output, resources = self.exporter.from_filename(
File "/usr/local/lib/python3.10/dist-packages/nbconvert/exporters/exporter.py", line 189, in from_filename
return self.from_file(f, resources=resources, **kw)
File "/usr/local/lib/python3.10/dist-packages/nbconvert/exporters/exporter.py", line 206, in from_file
return self.from_notebook_node(
File "/usr/local/lib/python3.10/dist-packages/nbconvert/exporters/pdf.py", line 194, in from_notebook_node
self.run_latex(tex_file)
File "/usr/local/lib/python3.10/dist-packages/nbconvert/exporters/pdf.py", line 164, in run_latex
return self.run_command(
File "/usr/local/lib/python3.10/dist-packages/nbconvert/exporters/pdf.py", line 139, in run_command
out, _ = p.communicate()
File "/usr/lib/python3.10/subprocess.py", line 1141, in communicate
stdout = self.stdout.read()
KeyboardInterrupt
^C
!jupyter nbconvert Copy_of_fin_analytics_try_2_j_comp.ipynb --to html
[NbConvertApp] Converting notebook Copy_of_fin_analytics_try_2_j_comp.ipynb to html
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/nbformat/reader.py", line 19, in parse_json
nb_dict = json.loads(s, **kwargs)
File "/usr/lib/python3.10/json/__init__.py", line 346, in loads
return _default_decoder.decode(s)
File "/usr/lib/python3.10/json/decoder.py", line 340, in decode
raise JSONDecodeError("Extra data", s, end)
json.decoder.JSONDecodeError: Extra data: line 3506 column 2 (char 3906009)
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/usr/local/bin/jupyter-nbconvert", line 8, in <module>
sys.exit(main())
File "/usr/local/lib/python3.10/dist-packages/jupyter_core/application.py", line 283, in launch_instance
super().launch_instance(argv=argv, **kwargs)
File "/usr/local/lib/python3.10/dist-packages/traitlets/config/application.py", line 992, in launch_instance
app.start()
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 423, in start
self.convert_notebooks()
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 597, in convert_notebooks
self.convert_single_notebook(notebook_filename)
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 560, in convert_single_notebook
output, resources = self.export_single_notebook(
File "/usr/local/lib/python3.10/dist-packages/nbconvert/nbconvertapp.py", line 488, in export_single_notebook
output, resources = self.exporter.from_filename(
File "/usr/local/lib/python3.10/dist-packages/nbconvert/exporters/exporter.py", line 189, in from_filename
return self.from_file(f, resources=resources, **kw)
File "/usr/local/lib/python3.10/dist-packages/nbconvert/exporters/exporter.py", line 207, in from_file
nbformat.read(file_stream, as_version=4), resources=resources, **kw
File "/usr/local/lib/python3.10/dist-packages/nbformat/__init__.py", line 174, in read
return reads(buf, as_version, capture_validation_error, **kwargs)
File "/usr/local/lib/python3.10/dist-packages/nbformat/__init__.py", line 92, in reads
nb = reader.reads(s, **kwargs)
File "/usr/local/lib/python3.10/dist-packages/nbformat/reader.py", line 75, in reads
nb_dict = parse_json(s, **kwargs)
File "/usr/local/lib/python3.10/dist-packages/nbformat/reader.py", line 25, in parse_json
raise NotJSONError(message) from e
nbformat.reader.NotJSONError: Notebook does not appear to be JSON: '{\n "cells": [\n {\n "cell_typ...